Webex
14:00-16:00 CET
Machine learning, neural networks and large-language models are swiftly being adopted in economic research. Currently, the use of generative AI is actively explored and embraced by economist in academia, central banks, think tanks and the financial sector. This workshop brings together cutting-edge research in two specific areas: first, the solving and estimation of economic models and, second, business cycle and inflation forecasting.
Organizing committee: Ernest Gnan – SUERF, Matthieu Darracq-Paries, Hanno Kase – ECB, Juha Kilponen, Esa Jokivuolle, Fabio Verona – Bank of Finland, Alessandro Notarpietro – Banca d’ Italia
Estimating nonlinear heterogeneous agents models with neural networks
Hanno Kase, ECB Presentation (pdf)Co-authors: Leonardo Melosi, University of Warwick, FRB Chicago, DNB, CEPR, and Matthias Rottner, Deutsche Bundesbank
Solving life-cycle models with rich asset structure using deep learning
Marlon Azinovic, University of Pennsylvania Presentation (pdf)Multi-agent deep reinforcement learning in macroeconomic modelling
Tohid Atashbar, IMF Presentation (pdf)Maximally Forward-Looking Core Inflation
Philippe Goulet Coulombe, Université du Québec à Montréal Presentation (pdf)Co-authors: Karin Klieber, Christophe Barrette, Maximilian Göbel
Exchange rate narratives
Kim Ristolainen, University of Turku Presentation (pdf)Co-author: Vito Cormun, Santa Clara University
